new england

The
journal of medicine
established in 1812 april 27, 2006 vol. 354 no. 17

Lung Recruitment in Patients

with the Acute Respiratory Distress Syndrome
Luciano Gattinoni, M.D., F.R.C.P., Pietro Caironi, M.D., Massimo Cressoni, M.D., Davide Chiumello, M.D.,
V. Marco Ranieri, M.D., Michael Quintel, M.D., Ph.D., Sebastiano Russo, M.D., Nicolò Patroniti, M.D.,
Rodrigo Cornejo, M.D., and Guillermo Bugedo, M.D.

A bs t r ac t

Background
In the acute respiratory distress syndrome (ARDS), positive end-expiratory pressure From the Istituto di Anestesiologia e Rian-
(PEEP) may decrease ventilator-induced lung injury by keeping lung regions open imazione, Fondazione Istituto di Ricove-
ro e Cura a Carattere Scientifico, Ospe-
that otherwise would be collapsed. Since the effects of PEEP probably depend on dale Maggiore Policlinico, Mangiagalli,
the recruitability of lung tissue, we conducted a study to examine the relationship Regina Elena di Milano, Università degli
between the percentage of potentially recruitable lung, as indicated by computed Studi di Milano, Milan (L.G., P.C., M.C.,
D.C.); the Dipartimento di Anestesia,
tomography (CT), and the clinical and physiological effects of PEEP. Azienda Ospedaliera San Giovanni Bat-
tista–Molinette, Università degli Studi di
Methods Torino, Turin, Italy (V.M.R.); Anaesthesi-
Sixty-eight patients with acute lung injury or ARDS underwent whole-lung CT dur- ologie II, Operative Intensivmedizin, Uni-
versitatsklinikum Gottingen, Gottingen,
ing breath-holding sessions at airway pressures of 5, 15, and 45 cm of water. The Germany (M.Q., S.R.); the Dipartimento di
percentage of potentially recruitable lung was defined as the proportion of lung Medicina Perioperatoria e Terapia Intensi-
tissue in which aeration was restored at airway pressures between 5 and 45 cm of va, Azienda Ospedaliera San Gerardo di
Monza, Università degli Studi Milano–
water. Bicocca, Milan (N.P.); and the Departa-
mentos de Anestesiologia y Medicina In-
Results tensiva, Facultad de Medicina, Pontificia
The percentage of potentially recruitable lung varied widely in the population, ac- Universidad Catolica de Chile, Santiago,
Chile (R.C., G.B.). Address reprint requests
counting for a mean (±SD) of 13±11 percent of the lung weight, and was highly cor- to Prof. Gattinoni at the Istituto di Anes-
related with the percentage of lung tissue in which aeration was maintained after tesiologia e Rianimazione, Fondazione
the application of PEEP (r2 = 0.72, P<0.001). On average, 24 percent of the lung could IRCCS–Ospedale Maggiore Policlinico,
Mangiagalli, Regina Elena di Milano, Uni-
not be recruited. Patients with a higher percentage of potentially recruitable lung versità degli Studi di Milano, Via F. Sforza
(greater than the median value of 9 percent) had greater total lung weights (P<0.001), 35, Milan 20122, Italy, or at gattinon@
poorer oxygenation (defined as a ratio of partial pressure of arterial oxygen to frac- policlinico.mi.it.
tion of inspired oxygen) (P<0.001) and respiratory-system compliance (P = 0.002), N Engl J Med 2006;354:1775-86.
higher levels of dead space (P = 0.002), and higher rates of death (P = 0.02) than pa- Copyright © 2006 Massachusetts Medical Society.
tients with a lower percentage of potentially recruitable lung. The combined physi-
ological variables predicted, with a sensitivity of 71 percent and a specificity of 59
percent, whether a patient’s proportion of potentially recruitable lung was higher or
lower than the median.
Conclusions
In ARDS, the percentage of potentially recruitable lung is extremely variable and is
strongly associated with the response to PEEP.

n engl j med 354;17 www.nejm.org april 27, 2006 1775

The New England Journal of Medicine

Downloaded from nejm.org on August 15, 2014. For personal use only. No other uses without permission.
Copyright © 2006 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

A
cute respiratory distress syn-
drome (ARDS) is a clinical syndrome char-
acterized by inflammatory pulmonary ede-
ma, severe hypoxemia, stiff lungs, and diffuse
endothelial and epithelial injury.1,2 Mechanical
ventilation is often implemented in these patients
to restore adequate oxygenation. However, it has
become evident over the past two decades that
mechanical ventilation itself can augment or cause
pulmonary damage that is indistinguishable from
that caused by ARDS.3 As a consequence, the ther-
apeutic target of mechanical ventilation in patients
with ARDS has shifted from the maintenance of
“normal gas exchange”4 to the protection of the
lung from ventilator-induced lung injury.5-7
The lung-protection strategy combines the use
of higher levels of positive end-expiratory pres-
sure (PEEP) (greater than 12 to 15 cm of water)
and low tidal volumes to prevent regional and
global stress and strain on the lung parenchy-
ma.8-10 Ventilation at low tidal volumes alone has
been shown to increase survival among patients
with acute lung injury or ARDS,11 and the addi-
tion of higher PEEP to low tidal volumes did not
further increase survival.12 In patients with low
levels of recruitable lung (i.e., lung tissue in which
aeration can be restored),13-16 however, the ap-
plication of higher levels of PEEP may be more
harmful than beneficial, since it will serve only to
increase inflation of lung regions that are already
open, increasing the stress and strain on these
regions.17 It follows that knowledge of the capac-
ity of the lung to become and remain recruited
should be a prerequisite for a rational determina-
tion of the levels of PEEP to be applied.
Using computed tomography (CT) to analyze
the entire lung in patients with ARDS, we mea-
sured the percentage of lung that can be re-
cruited, termed “potentially recruitable lung,” by
increasing airway pressures.18,19 We also investi-
gated the relationship between the percentage of
lung that can be recruited by this maneuver and
the changes in physiological respiratory variables
during mechanical ventilation with lower or
higher PEEP.
Me thods
patients
The patients were studied from June 2003 through to the values used in everyday clinical treatment.
January 2005 at four university hospitals. The study After 20 minutes, the systemic arterial and cen-
was approved by the institutional review board of tral venous pressures and blood gas tensions, min-
each hospital, and written informed consent was
obtained according to the national regulations of
the participating institutions (consent was delayed
in Italy until after the patients had recovered from
the effects of sedation, obtained from a legal rep-
resentative in Germany, and obtained from the next
of kin in Chile; for details see the Supplementary
Appendix, available with the full text of this article
at www.nejm.org).
Patients were enrolled if they met the standard
criteria for acute lung injury: a ratio of the partial
pressure of arterial oxygen to the fraction of in-
spired oxygen (PaO2:FiO2) of less than 300, the
presence of bilateral pulmonary infiltrates on the
chest radiograph, and no clinical evidence of left
atrial hypertension (defined by a pulmonary-cap-
illary wedge pressure of 18 mm Hg or less, if
measured).20 The exclusion criteria were an age
of less than 16 years, pregnancy, and chronic ob-
structive pulmonary disease, according to the pa-
tient’s medical history. The underlying cause of
acute lung injury or ARDS was recorded by each
institution, but no specific classifications were
defined a priori. Patients with healthy lungs and
patients with unilateral pneumonia who under-
went CT for clinical purposes from April 2001
through June 2005 were retrospectively selected
from five hospitals and included in the study for
comparison (Fig. 1 and the Supplementary Ap-
pendix).
peep Trial
The clinical characteristics of the patients, respira-
tory variables, and ventilator settings were record-
ed before the study. Immediately before each step
of the PEEP trial, as well as before each CT session,
a recruitment maneuver — that is, a sustained in-
flation of the lungs to higher airway pressures and
volumes than are obtained during tidal ventilation
— was performed in which the patient underwent
ventilation for two minutes in the pressure-con-
trolled mode at an inspiratory plateau pressure of
45 cm of water, a PEEP of 5 cm of water, a respira-
tory rate of 10 breaths per minute, and a 1:1 ratio
of inspiration to expiration.21,22 After the recruit-
ment maneuver, PEEP at a level of 5 or 15 cm of
water was randomly applied (Fig. 1). The tidal vol-
ume (8 to 10 ml per kilogram of predicted body
weight), FiO2, and respiratory rate were identical

1776 n engl j med 354;17 www.nejm.org april 27, 2006

The New England Journal of Medicine

Downloaded from nejm.org on August 15, 2014. For personal use only. No other uses without permission.
Copyright © 2006 Massachusetts Medical Society. All rights reserved.
 lung recruitment in patients with the acute respir atory distress syndrome

Study group Comparison groups

68 Patients enrolled with Retrospective selection

diagnosis of acute from hospital database
lung injury or ARDS of 63 patients with lung
injury other than acute
lung injury or ARDS who
underwent whole-lung CT
for diagnostic purposes
Recruitment maneuvers

PEEP trial with PEEP

at 5 and 15 cm of water 1 With congestive heart
(random sequence) and failure excluded
physiological variables

Recruitment maneuver

CT with inspiratory plateau

62 With pneumonia
pressure of 45 cm of water

Recruitment maneuvers

CT with PEEP at 5 and 15 cm 28 With bilateral pneumonia

of water (random sequence) excluded

Quantitative analysis Quantitative analysis

34 Had unilateral
34 With lower percentage 34 With higher percentage 39 Had healthy lungs
pneumonia

Figure 1. Enrollment and Study Protocol.

In the study group, a recruitment maneuver was performed immediately before application of each PEEP level. In the comparison
groups, patients with bilateral pneumonia were excluded from the analysis to limit the possible confounding factors caused by the par-
tial overlapping between patients with less severe acute lung injury or ARDS and patients with bilateral pneumonia (see the Supplemen-
tary Appendix for further details). Therefore, only patients with unilateral pneumonia, who by definition did not meet the inclusion crite-
ria for acute lung injury or ARDS, were included. The group with a lower percentage of potentially recruitable lung includes patients with
potentially recruitable lung values at or below the overall median of 9 percent, and the group with a higher percentage of potentially re-
cruitable lung includes patients with values above the median.

ute ventilation, and inspiratory plateau pressure Standard formulas were used to calculate the
were recorded. The dead-space fraction and the right-to-left intrapulmonary shunt fraction, alve-
end-tidal partial pressure of carbon dioxide were olar dead-space fraction, and respiratory-system
measured with a CO2SMO monitor (Novametrix). compliance (see the Supplementary Appendix).

n engl j med 354;17 www.nejm.org april 27, 2006 1777

The New England Journal of Medicine

Downloaded from nejm.org on August 15, 2014. For personal use only. No other uses without permission.
Copyright © 2006 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l
Computed Tomography
The CT scanner was set as follows: collimation,
5 mm; interval, 5 mm; bed speed, 15 mm per sec-
ond; voltage, 140 kV; and current, 240 mA. A whole-
lung CT scan was performed at an inspiratory-
plateau pressure of 45 cm of water during an
end-inspiratory pause (ranging from 15 to 25 sec-
onds) and thereafter at PEEP values of 5 and 15
cm of water applied in a random order during an
end-expiratory pause (ranging from 15 to 25 sec-
onds). Immediately before each CT scan was ob-
tained, a recruitment maneuver was performed,
as described above (Fig. 1); the ventilator settings
were otherwise kept identical to those used dur-
ing the PEEP trial. The patients included in the
comparison groups underwent only one CT of the
whole lung, for diagnostic purposes. The cross-
sectional lung images were processed and ana-
lyzed by a custom-designed software package, as
described previously19 (see the Supplementary Ap-
pendix). Briefly, the outline of the lungs was man-
ually drawn in each image, excluding the hilar
vessels, by investigators unaware of the airway
pressure applied. Specific lung weight was as-
sumed to be equal to 1, and the total lung weight
was calculated from the physical density of the
lung expressed in Hounsfield units. Similarly, the
tissue weights of lung regions with different de-
grees of aeration were calculated. The regions were
classified as nonaerated (density between +100 and
–100 Hounsfield units), poorly aerated (density
between –101 and –500 Hounsfield units), normal-
ly aerated (density between –501 and –900 Houns-
field units), and hyperinflated (density between
–901 and –1000 Hounsfield units). The percentage
of potentially recruitable lung was defined as the
proportion of the total lung weight accounted for
by nonaerated lung tissue in which aeration was
restored (according to CT) by an airway pressure
of 45 cm of water from an airway pressure of 5 cm
of water.
Statistical Analysis
Comparison of prestudy clinical variables, respi-
ratory physiological variables, and CT results was
performed by one-way analysis of variance or
Student’s t-test in the case of variables that were
normally distributed; by the Kruskal–Wallis test,
the Wilcoxon test, or two-way analysis of variance
on a rank-sum test in the case of variables that
did not appear normally distributed on graphic
inspection; and by the chi-square test or Fisher’s
exact test in the case of qualitative variables. When
1778 n engl j med 354;17
The New England Journal of Medicine
Downloaded from nejm.org on August 15, 2014. For personal use only. No other uses without permission.
Copyright © 2006 Massachusetts Medical Society. All rights reserved.
of m e dic i n e
analysis of variance revealed a significant differ-
ence, Bonferroni’s t-test or Dunn’s test was used,
as appropriate, to correct for multiple compari-
sons. Mortality rates were analyzed by the chi-
square test. Mortality rates were based on the
number of deaths occurring in the intensive care
unit (ICU) among patients with acute lung injury
or ARDS and the number of deaths occurring in
the hospital among patients with unilateral
pneumonia. Multiple backward logistic-regres-
sion analysis was used to investigate the possible
association between outcome and the percentage
of potentially recruitable lung, as well as other
measurements used to estimate the severity of
the systemic illness and of the lung injury. The
Hosmer–Lemeshow goodness-of-fit test and the
C statistic were used to verify the adequacy of
the models.
To obtain a bedside estimate of the percentage
of potentially recruitable lung using only physi-
ological respiratory measurements, we measured
the changes in the PaO2:FiO2, the partial pres-
sure of arterial carbon dioxide (PaCO2), the per-
centage of alveolar dead space, and respiratory-
system compliance associated with increasing the
PEEP from 5 to 15 cm of water while minute ven-
tilation and FiO2 were held constant. An increase
in the PaO2:FiO2, a decrease in the PaCO2 or al-
veolar dead space, or an increase in respiratory-
system compliance was defined as a positive re-
sponse, and any change in the opposite direction
was defined as a negative response, irrespective
of the magnitude of the change. P values of less
than 0.05 were considered to indicate statistical
significance. All reported P values are two-sided.
Data are expressed as means (±SD) and 95 per-
cent confidence intervals when appropriate.
R e sult s
A total of 68 patients were enrolled in the study:
19 had acute lung injury without ARDS, and 49
had ARDS (Fig. 1 and Table 1). The overall mor-
tality rate in the ICU among the study popula-
tion was 28 percent. The percentage of poten-
tially recruitable lung, as assessed by CT, varied
widely within the study population (Fig. 2); the
average was 13±11 percent of the total lung weight
(95 percent confidence interval, 10 to 16 percent;
median, 9 percent), corresponding to an absolute
weight of 217±232 g of recruitable lung tissue
(95 percent confidence interval, 161 to 273; me-
dian, 134).
www.nejm.org april 27, 2006
 lung recruitment in patients with the acute respir atory distress syndrome

Table 1. Baseline Characteristics of the Study Population.*

Patients with Lower Patients with Higher

Overall Percentage of Percentage of
Population Potentially Recruitable Potentially Recruitable
Characteristic (N = 68) Lung (N = 34)† Lung (N = 34)† P Value‡
Age — yr 55±17 56±16 53±18 0.48
Female sex — no. (%) 33 (49) 15 (44) 18 (53) 0.47
Body-mass index 25±5 26±5 24±4 0.21
SAPS II score§ 37±11 37±12 36±9 0.91
Tidal volume — ml/kg of predicted body 8.8±1.9 8.9±2.0 8.8±1.7 0.78
weight
Minute ventilation — liters/min 9.8±3.0 9.5±2.7 10.1±3.3 0.45
Respiratory rate — breaths/min 18±7 17±6 19±7 0.57
PEEP — cm of water 11.1±3.0 10.8±2.9 11.5±3.1 0.34
Plateau pressure — cm of water 25±4 23±3 26±4 0.005
Respiratory-system compliance 44±17 49±16 40±18 0.02
— ml/cm of water¶
PaO2:FiO2 200±77 225±70 176±77 0.008
FiO2 0.50±15 0.46±10 0.54±18 0.07
PaCO2 — mm Hg 42±14 38±8 46±17 0.04
Arterial pH 7.40±0.08 7.41±0.08 7.37±0.07 0.01
Cause of lung injury — no. (%)
Pneumonia 25 (37) 7 (21) 18 (53) 0.01
Sepsis 24 (35) 17 (50) 7 (21) 0.02
Aspiration 4 (6) 3 (9) 1 (3) 0.61
Trauma 3 (4) 3 (9) 0 0.24
Other∥ 12 (18) 4 (12) 8 (24) 0.34
Fluid balance before study — ml/day** 1413±2027 1427±2016 1398±2071 0.97
Days of ventilation before study†† 5±6 5±6 6±6 0.50
Type of lung injury 0.02
Acute lung injury 19 14 5
ARDS 49 20 29

* Plus–minus values are means ±SD. Because of rounding, percentages may not total 100. The body-mass index is the
weight in kilograms divided by the square of the height in meters.
† Patients in the group with a lower percentage of potentially recruitable lung had values at or below 9 percent, the me-
dian value for the study population, and patients in the group with a higher percentage of potentially recruitable lung
had values greater than 9 percent.
‡ P values were obtained by Student’s t-test, Wilcoxon’s test, Fisher’s exact test, or the chi-square test, as appropriate.
§ The Simplified Acute Physiology Score (SAPS II)23 was used to assess the severity of systemic illness at study entry.
Scores can range from 0 to 163, with higher scores indicating more severe illness.
¶ Respiratory-system compliance was calculated as the ratio of the tidal volume to the difference between inspiratory
plateau pressure and PEEP.
∥ Other causes of acute lung injury included anaphylactic shock, recent surgery, and bone marrow transplantation.
** The fluid balance before the study was the average daily fluid balance for each patient during the last five days before
the study.
†† Days of mechanical ventilation before the study were counted from the day of intubation (day 0) to the beginning of
the study.

Functional Anatomy According to CT able lung (Fig. 3A); the average values were 2±4 per-
Findings and Response to PEEP cent of total lung weight in quartile 1 (range, −9.2
The study population was divided into quartiles to 5.7 percent), 7±1 percent in quartile 2 (range,
according to the percentage of potentially recruit- 5.8 to 9.4 percent), 14±3 percent in quartile 3 (range,

n engl j med 354;17 www.nejm.org april 27, 2006 1779

The New England Journal of Medicine

Downloaded from nejm.org on August 15, 2014. For personal use only. No other uses without permission.
Copyright © 2006 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

A Figure 2. Frequency Distribution of Patients According

Average Average to the Percentage of Potentially Recruitable Lung (Panel A)
Lung Weight, Lung Weight, and CT Images at Airway Pressures of 5 and 45 cm of
59±51 g 374±236 g Water from Patients with a Lower Percentage of Poten-
tially Recruitable Lung (Panel B) and Those with a
Lower Higher Higher Percentage of Potentially Recruitable Lung
24 (Panel C).
22 Patients with acute lung
Panel A shows the frequency distribution of the 68 pa-
5±4% 21±10% tients in the overall study group according to the per-
20 injury without ARDS
Patients with ARDS centage of potentially recruitable lung, expressed as
18
the percentage of total lung weight. Acute lung injury
16
No. of Patients

without ARDS was defined by a PaO2:FiO2 of less than

14 300 but not less than 200, and ARDS was defined by a
12 PaO2:FiO2 of less than 200. The percentage of poten-
10 tially recruitable lung was defined as the proportion of
8 lung tissue in which aeration is restored at airway pres-
6 sures between 5 and 45 cm of water. Panel B shows
4 representative CT slices of the lung obtained 2 cm
2 above the diaphragm dome at airway pressures of 5 cm
of water (left) and 45 cm of water (right) from a patient
0
with a lower percentage of potentially recruitable lung
0 0
5 5
¡5 ¡5

10 10
15 15
20 20
25 25
30 30
35 35
40 40
45 45
50 50
55 55
60 60

75 75
80
65 65
70 70
to
to

(at or below the median value of 9 percent of total lung

to

to
to
to
to
to
to
to
to
to
to
to

to
to
to
to
0
¡1

weight). Lung injury developed in the patient after an

Amount of Potentially Recruitable Lung (% total lung weight) episode of severe acute pancreatitis (PaO2:FiO2, 296 at
an airway pressure of 5 cm of water; PaCO2, 34 mm Hg;
B Lower Percentage of Potentially Recruitable Lung and respiratory-system compliance, 44 ml per centime-
ter of water). The percentage of potentially recruitable
5 cm of water 45 cm of water
lung was 4 percent, and the proportion of consolidated
lung tissue was 33 percent of the total lung weight.
Panel C shows representative CT slices of the lung ob-
tained 2 cm above the diaphragm dome at airway pres-
sures of 5 cm of water (left) and 45 cm of water (right)
from a patient in the group with a higher percentage of
potentially recruitable lung. Lung injury developed in
the patient after an episode of severe pneumonia
(PaO2:FiO2, 106 at a PEEP of 5 cm of water; PaCO2,
C Higher Percentage of Potentially Recruitable Lung 58 mm Hg; and respiratory-system compliance, 25 ml
per cm of water). The percentage of potentially recruit-
5 cm of water 45 cm of water able lung was 37 percent, and the proportion of con-
solidated lung tissue was 27 percent of the total lung
weight.

The decrease in the percentage of nonaerated

9.5 to 18.6 percent), and 28±10 percent in quar-
tile 4 (range, 18.7 to 59.3 percent). In each group,
the increase in airway pressure from 5 to 15 to 45
cm of water induced a progressive increase in the
percentage of hyperinflated and normally aerat-
ed lung tissue (P<0.01 for both variables), paral-
leled by a decrease in the percentage of nonaer-
ated lung tissue (P<0.01). In contrast, in all four
groups, about 24 percent of the lung could not be
recruited, even at an airway pressure of 45 cm of
water.
lung tissue as PEEP was raised from 5 to 15 cm of
water was highly correlated with the percentage
of potentially recruitable lung (r2 = 0.72, P<0.001)
(Fig. 3B). The near-constant fraction of the per-
centage of potentially recruitable lung that re-
mained recruited at a PEEP of 15 cm of water was
about 50 percent, irrespective of its absolute per-
centage, as indicated by the slope of the plot in
Figure 3B.
Clinical Characteristics and Overall Severity
of Lung Injury
We divided the patients into two groups accord-
ing to the percentage of potentially recruitable
lung: at or below the median value of 9 percent of
total lung weight or greater than the median value

1780 n engl j med 354;17 www.nejm.org april 27, 2006

The New England Journal of Medicine

Downloaded from nejm.org on August 15, 2014. For personal use only. No other uses without permission.
Copyright © 2006 Massachusetts Medical Society. All rights reserved.
 lung recruitment in patients with the acute respir atory distress syndrome

Figure 3. Lung Recruitment in Response to Changes A

in Airway Pressure in Patients with Acute Lung Injury Potentially recruitable Hyperinflated
or ARDS, According to the Percentage of Potentially Consolidated Normally aerated
Recruitable Lung. Poorly aerated
Panel A shows the proportion of total lung tissue clas- Lung Weight, Lung Weight, Lung Weight, Lung Weight,
sified as nonaerated, poorly aerated, normally aerated, 1233±333 g 1298±327 g 1630±458 g 1840±619 g*
¶ ¶ ¶ ‡¶

Tissue Weight (% of total lung weight)

and hyperinflated in response to three values of airway 100 †¶
§¿ ¶ ‡¶
pressure in patients with different percentages of po- 90
¿ Aerated
tentially recruitable lung. The study population was di- 80
vided into quartiles of 17 patients each according to 70 §
the percentage of potentially recruitable lung. Nonaer- 60 ¶
ated lung tissue was also divided into potentially re- 50
¶
†¶
cruitable tissue (nonaerated tissue in which aeration Nonaerated
40
was restored at airway pressures between 5 and 45 cm †¶
30 §
of water) and consolidated tissue (tissue remaining †¶
20
nonaerated despite an airway pressure of 45 cm of wa-
10
ter). The asterisk denotes P<0.01 for the comparison
0
with patients with a very low or low percentage of po-

wa r
r
cm f w er

cm f w er
wa r
r

cm f w er
wa r
r

cm f w er
wa r
r
of ate
te

of ate
te

of ate
te

of ate
te
tentially recruitable lung (first and second quartiles),

45 m o wat

45 o at

45 o at

45 o at
cm f w

cm f w

cm f w
c f
the daggers P<0.01 for the comparison with patients in

15 m o

15 o

15 o

15 o
cm

cm

cm
c
the other quartiles, the double dagger P<0.01 for the

5
comparison with patients with a very low percentage of Quartile 1 Quartile 2 Quartile 3 Quartile 4
potentially recruitable lung (first quartile), the section ( 9.2 to 5.7%) (5.8 to 9.4%) (9.5 to 18.6%) (18.7 to 59.3%)
marks P<0.05 for the comparison with an airway pres-
sure of 45 cm of water in patients within the same B
quartile, the paragraph mark P<0.01 for the compari-
Lower percentage of potentially recruitable lung (N= 34)
son with airway pressures of 15 and 45 cm of water for
Higher percentage of potentially recruitable lung (N= 34)
patients within the same quartile, and the double
slashes P<0.05 for the comparison with a PEEP value 35
of 15 cm of water for patients within the same quartile.
30
Panel B shows lung recruitment induced by increasing
Lung Recruitment between PEEP
Values of 5 and 15 cm of Water

PEEP from 5 to 15 cm of water in the overall study pop- 25

(% of total lung weight)

ulation — that is, the decrease in nonaerated lung tis-

sue between PEEP values of 5 and 15 cm of water, as a 20
function of the percentage of potentially recruitable
15
lung; both values are expressed as proportions of the
total lung weight measured at a baseline PEEP of 5 cm 10
of water (r 2 = 0.72, P<0.001, slope = 0.52 and y inter-
cept = 1.03). A linear function (y = ax + y 0) was used. 5

for the study population. In the prestudy period, ¡5

the two groups had similar clinical characteris- ¡10 0 10 20 30 40 50 60 70

tics with regard to age, severity of illness (as as-
sessed by the Simplified Acute Physiology Score
[SAPS II]23), daily fluid balance, and number of
days of mechanical ventilation before the begin-
ning of the study (Table 1). The tidal volume and
PEEP level used clinically for mechanical ventila-
tion were similar in the two groups. In contrast,
at baseline, patients in the group with a higher
percentage of potentially recruitable lung had
a lower PaO2:FiO2 (P = 0.008), a higher PaCO2
(P =0.04), and lower respiratory-system compli-
ance (P = 0.02) than those in the group with a
lower percentage of potentially recruitable lung
(Table 1). Acute lung injury or ARDS resulting
from sepsis was more frequent among patients
Percentage of Potentially Recruitable Lung
(% of total lung weight)
in the group with a lower percentage of recruit-
able lung (P = 0.02), whereas acute lung injury or
ARDS resulting from pneumonia was more fre-
quent among patients in the group with a higher
percentage (P = 0.01) (Table 1 and the Supplemen-
tary Appendix).
The association between the percentage of
potentially recruitable lung and the severity of the
overall lung injury was examined at a PEEP of
5 cm of water. The total lung weight was greater
(P<0.001), the proportion of nonaerated lung
tissue was higher (P = 0.001), the PaO2:FiO2 was

n engl j med 354;17 www.nejm.org april 27, 2006 1781

The New England Journal of Medicine

Downloaded from nejm.org on August 15, 2014. For personal use only. No other uses without permission.
Copyright © 2006 Massachusetts Medical Society. All rights reserved.
 1782
Table 2. Baseline Characteristics, Functional Anatomy According to CT Findings, and Mortality Rates in Patients with Healthy Lungs, Patients with Unilateral Pneumonia,
and Patients with Acute Lung Injury or ARDS and with Lower or Higher Percentages of Potentially Recruitable Lung.*

Patients with Healthy Patients with Unilateral P

Variable Lungs (N = 39) Pneumonia (N = 34) Patients with Acute Lung Injury or ARDS† Value‡
Overall Lower Percentage of Higher Percentage of
Population Potentially Recruitable Potentially Recruitable
(N = 68) Lung (N = 34) Lung (N = 34)

Age — yr 62±19 65±18 55±17§ 56±16 53±18 0.01

The

SAPS II score¶ — 35±16 37±11 37±12 36±9 0.18

Total lung weight — g 850±201 1215±329∥** 1500±506∥†† 1266±327** 1735±547‡‡ <0.001

Nonaerated lung tissue — % of total lung weight§§ 3±2 28±14∥** 37±16∥†† 30±12** 44±17‡‡ <0.001

Aerated lung tissue — % of total lung weight§§ 97±2 72±14∥** 63±16∥†† 70±12** 56±17‡‡ <0.001

n engl j med 354;17

PaO2:FiO2¶¶ — 219±103 165±69 194±65** 135±60‡‡ 0.01

PaCO2 — mm Hg ¶¶ — 40±8 42±9 39±7 44±10∥∥ 0.50

Respiratory-system compliance — ml/cm of water*** — — 44±19 51±19 38±15 0.002

www.nejm.org
n e w e ng l a n d j o u r na l

The New England Journal of Medicine

Dead space — % of tidal volume††† — — 57±13 51±12 63±13 0.002
of

Shunt — % of cardiac output‡‡‡ — — 39±15 34±12 45±17 0.008

Mortality — no. (%) of patients§§§ — 6 (18) 19 (28) 5 (15) 14 (41)¶¶¶ 0.25

april 27, 2006

Copyright © 2006 Massachusetts Medical Society. All rights reserved.

* Two statistical analyses are reported: the comparison between patients with healthy lungs, patients with unilateral pneumonia, and the overall population of patients with acute lung
m e dic i n e

injury or ARDS, as well as the comparison between each single group of patients (patients with healthy lungs, patients with unilateral pneumonia, patients with acute lung injury or
ARDS with a lower percentage of potentially recruitable lung, and patients with acute lung injury or ARDS with a higher percentage of potentially recruitable lung). Because of round-
ing, percentages may not total 100. Plus–minus values are means ±SD.
† Patients in the group with a lower percentage of potentially recruitable lung had values at or below 9 percent, the median value for the study population, and patients in the group
with a higher percentage of potentially recruitable lung had values greater than 9 percent.

Downloaded from nejm.org on August 15, 2014. For personal use only. No other uses without permission.
‡ P values were obtained by one-way analysis of variance, the Kruskal–Wallis test, Student’s t-test, Wilcoxon’s test, and the chi-square test, as appropriate; comparisons were made be-
tween patients with healthy lungs, patients with unilateral pneumonia, and the overall population of patients with acute lung injury or ARDS.
§ P<0.05 for the comparison with patients with unilateral pneumonia.
¶ The Simplified Acute Physiology Score (SAPS II)23 was used to assess the severity of systemic illness at study entry. Scores can range from 0 to 163, with higher scores indicating
more severe illness.
∥ P<0.01 for the comparison with patients with healthy lungs.
 ** P<0.05 for the comparison with patients with healthy lungs; comparisons were made between patients with healthy lungs, patients with unilateral pneumonia, patients with acute
lung injury or ARDS and a lower percentage of potentially recruitable lung, and patients with acute lung injury or ARDS and a higher percentage of potentially recruitable lung.
†† P<0.01 for the comparison with patients with either healthy lungs or unilateral pneumonia.
‡‡ P<0.001 for the comparison with other groups of patients; comparisons were made between patients with healthy lungs, patients with unilateral pneumonia, patients with acute
lung injury or ARDS and a lower percentage of potentially recruitable lung, and patients with acute lung injury or ARDS and a higher percentage of potentially recruitable lung.
§§ Nonaerated lung tissue is the portion of lung parenchyma with a density between +100 and –100 Hounsfield units, and aerated lung tissue is the portion of lung parenchyma with a
density between –101 and –1000 Hounsfield units. For simplicity, normally aerated, poorly aerated, and hyperinflated lung tissue were classified as aerated tissue. Thus, the percent-
age of aerated tissue is the percentage of lung tissue open to aeration, irrespective of the specific degree of aeration.
¶¶ Measurements of PaO2:FiO2 and PaCO2 were available for 31 patients with unilateral pneumonia.
∥∥ P<0.001 for the comparison with patients with acute lung injury or ARDS and a lower percentage of potentially recruitable lung.
*** Respiratory-system compliance was calculated as the ratio of the tidal volume to the difference between inspiratory plateau pressure and PEEP.
††† Dead space was calculated by a standard formula (see the Supplementary Appendix). This measurement, obtained at a PEEP value of 5 cm of water, was available for 48 patients
with acute lung injury or ARDS (23 in the group with a lower percentage of recruitable lung and 25 in the group with a higher percentage of recruitable lung).
‡‡‡ The intrapulmonary right-to-left shunt was calculated by a standard formula (see the Supplementary Appendix). This measurement, obtained at a PEEP value of 5 cm of water, was
available for 60 patients (29 in the group with a lower percentage of recruitable lung and 31 in the group with a higher percentage of recruitable lung).
§§§ The number of deaths occurring in the hospital was recorded for patients with unilateral pneumonia and the number of deaths in the ICU for patients with acute lung injury or
ARDS. The mean length of stay after admission was 25±21 days for patients in the hospital and 29±27 days for patients in the ICU; length of stay includes both patients who died
and those who were discharged.
¶¶¶ P<0.05 for the comparison with patients with acute lung injury or ARDS and a lower percentage of potentially recruitable lung.

n engl j med 354;17

www.nejm.org
1.14), respectively.

The New England Journal of Medicine

april 27, 2006

Copyright © 2006 Massachusetts Medical Society. All rights reserved.

lung recruitment in patients with the acute respir atory distress syndrome

Downloaded from nejm.org on August 15, 2014. For personal use only. No other uses without permission.
centage of potentially recruitable lung (Table 2).
group with a higher percentage of potentially re-

pendently associated with an increased risk of

potentially recruitable lung appeared to be inde-
illness, and the percentage of potentially recruit-

higher percentage of recruitable lung. The func-

cific for acute lung injury or ARDS by retrospec-
of death (Fig. 4A). To investigate whether there
An association was observed between the per-

mechanical ventilation) (see the Supplementary

lungs and a group of 34 patients with unilateral
death (P = 0.47 by the Hosmer–Lemeshow good-
system compliance at a PEEP of 5 cm of water

patients with acute lung injury or ARDS was

nia, the total lung weight and the proportion of
able lung, the percentage of nonaerated lung tis-
and the mortality rate was higher (P = 0.02) in the
lower (P<0.001), the respiratory-system compli-

tween patients with unilateral pneumonia and

parisons). However, the greatest differences be-
acute lung injury or ARDS (P<0.01 for all com-
portion of normally aerated lung tissue was higher
with either healthy lungs or unilateral pneumo-
taneously and 14 patients who were undergoing
pneumonia (20 patients who were breathing spon-
tively evaluating a group of 39 patients with healthy
We further characterized these findings as spe-
percentage of potentially recruitable lung were
one-point increase in the SAPS II score and in the
analysis for independent predictors of mortality
(P = 0.002), the shunt fraction was higher (P = 0.008),

cruitable lung than in the group with a lower per-

tional anatomy, as determined by CT, appeared to

of the study population (Fig. 1). Among patients
Appendix) and comparing the findings with those

than among the overall population of patients with

injury. The SAPS II score and the percentage of
were included as markers of the severity of lung
sue, the PaO2:FiO2, the PaCO2, and the respiratory-
as a marker of the overall severity of the systemic
measurements of severity of illness, a multivariate
was an association between mortality and other
centage of potentially recruitable lung and the risk
ance was lower (P = 0.002), the PaCO2 was higher

among the patients in the latter group who had a

nonaerated lung tissue were lower and the pro-
was performed. The SAPS II score was included
(P = 0.02), the percentage of dead space was higher

and 1.08 (95 percent confidence interval, 1.01 to

1.08 (95 percent confidence interval, 1.02 to 1.15)
ness-of-fit test; C = 0.78); the odds ratios for each

1783
 The n e w e ng l a n d j o u r na l of m e dic i n e

A Figure 4. Mortality in Relation to the Percentage of

60 Potentially Recruitable Lung (Panel A) and Pulmonary
Anatomy According to CT Findings in Patients with
50
Healthy Lungs, Patients with Unilateral Pneumonia,
and Patients with Acute Lung Injury or ARDS (Panel B).
40
Panel A shows the mortality rate in the ICU (mean
Mortality (%)

length of stay, 29±27 days; range, 2 to 163) among

patients with acute lung injury or ARDS (P = 0.34 by
30
the Hosmer–Lemeshow goodness-of-fit test; C = 0.72).
Results are shown for quartiles of 17 patients each
20 according to the percentage of potentially recruitable
lung. Panel B shows the weights of total lung tissue
10 and nonaerated lung tissue in 39 patients with healthy
lungs, 34 patients with unilateral pneumonia, 34 pa-
0 tients with acute lung injury or ARDS with a lower per-
Quartile 1 Quartile 2 Quartile 3 Quartile 4 centage of potentially recruitable lung (at or below the
(¡9.2 to 5.7%) (5.8 to 9.4%) (9.5 to 18.6%) (18.7 to 59.3%) overall median value of 9 percent), and 34 patients
with acute lung injury or ARDS with a higher percent-
age of potentially recruitable lung. Data from the pa-
B
tients with healthy lungs and unilateral pneumonia
3500 Total lung tissue were obtained from a whole-lung CT obtained for di-
Nonaerated lung tissue
agnostic purposes. Data from patients with acute lung
3000
injury or ARDS were obtained from a whole-lung CT
Weight of Lung Tissue (g)

performed at a PEEP of 5 cm of water. Solid lines rep-

2500
resent mean values of total lung weight, and dashed
2000 lines mean values of nonaerated lung-tissue weight.
*†
Asterisks denote P<0.01 for the comparison with pa-
1500 tients with healthy lungs; daggers denote P<0.01 for
* * the comparison between the groups of patients with
1000 acute lung injury or ARDS and a higher percentage of
*† potentially recruitable lung and the other three groups.
500
* *

0 15 cm of water: an increase in the PaO2:FiO2, a

Patients Patients Patients Patients
with Healthy with Unilateral with Acute with Acute
decrease in the PaCO2, or an increase in the re-
Lungs Pneunomia Lung Injury Lung Injury spiratory-system compliance. However, the pow-
or ARDS or ARDS er of this test to predict which patients had a
and Lower and Higher
Percentage Percentage higher percentage of potentially recruitable lung
of Potentially of Potentially had a sensitivity of 71 percent and a specificity of
Recruitable Recruitable
Lung Lung 59 percent. A post hoc analysis was used to eval-
uate other combinations of different physiologi-
cal respiratory variables that were tested as pre-
be very similar in patients with unilateral pneu- dictors of the percentage of potentially recruitable
monia and patients with acute lung injury or lung. Among these combinations, a PaO2:FiO2 of
ARDS and a lower percentage of recruitable lung less than 150 at a PEEP of 5 cm of water had a
(Table 2 and Fig. 4B). sensitivity of 74 percent and a specificity of 79
percent. The combination of variables that yield-
Prediction of the Percentage of Potentially ed the best results appeared to be the presence of
Recruitable Lung at least two of the following: a PaO2:FiO2 of less
To provide a bedside estimate of the percentage of than 150 at a PEEP of 5 cm of water, any decrease
potentially recruitable lung, we initially hypoth- in alveolar dead space, and an increase in respi-
esized that in patients with a higher percentage ratory-system compliance when PEEP was in-
of potentially recruitable lung, at least two of the creased from 5 to 15 cm of water (sensitivity, 79
following three changes in respiratory variables percent; specificity, 81 percent) (see the Supple-
would occur when PEEP was increased from 5 to mentary Appendix).

1784 n engl j med 354;17 www.nejm.org april 27, 2006

The New England Journal of Medicine

Downloaded from nejm.org on August 15, 2014. For personal use only. No other uses without permission.
Copyright © 2006 Massachusetts Medical Society. All rights reserved.
 lung recruitment in patients with the acute respir atory distress syndrome
Dis cus sion
CT revealed that the percentage of potentially re-
cruitable lung varied widely among patients with
acute lung injury or ARDS, from a negligible frac-
tion to more than 50 percent of the total lung
weight. Furthermore, we demonstrated that the
effect of PEEP on lung recruitment was closely
associated with the percentage of potentially re-
cruitable lung and that the percentage of poten-
tially recruitable lung was itself highly correlated
with the overall severity of lung injury.
In clinical practice, lung recruitment is usually
considered a useful strategy.8,9,24 For this reason,
it has been suggested that the condition of pa-
tients with a high percentage of potentially re-
cruitable lung is better than that of patients with
a lower percentage of potentially recruitable lung,
given the presence of similar degrees of lung in-
jury. Surprisingly, among our patients, a higher
percentage of potentially recruitable lung corre-
lated with markedly poorer gas exchange and re-
spiratory mechanics, a greater severity of lung
injury, and a higher mortality rate, even though
the severity of their systemic illness at study en-
try, as assessed by the SAPS II score, was similar
in patients with higher and those with lower per-
centages of potentially recruitable lung (see the
Supplementary Appendix). An association between
the percentage of potentially recruitable lung
and the severity of lung injury, although unex-
pected, appears logical. In healthy lungs, the per-
centage of potentially recruitable lung is close to
0 percent, because the alveolar units are usually
not collapsed. When ARDS affects the lungs, the
extent of the inflammatory pulmonary edema is
linked to the likelihood of gravity-dependent al-
veolar collapse18,25 and thus to the percentage
of potentially recruitable lung. It is tempting to
speculate that the “core disease” is reflected by
the unrecruitable lung tissue at 45 cm of water
(about 24 percent of the total lung weight), where-
as the extent of the surrounding inflammatory
reaction26 is reflected by the collapsed but open-
able lung tissue — that is, the potentially recruit-
able lung.
The use of respiratory physiological variables
that can be measured at the bedside to ascertain
the percentage of potentially recruitable lung was
less specific and sensitive than expected. How-
ever, we think that analysis of CT findings can
identify the increase in aeration of previously col-
n engl j med 354;17
The New England Journal of Medicine
Downloaded from nejm.org on August 15, 2014. For personal use only. No other uses without permission.
Copyright © 2006 Massachusetts Medical Society. All rights reserved.
lapsed lung regions (“anatomical” lung recruit-
ment19), whereas changes in respiratory physiolog-
ical variables are specifically related to “functional”
recruitment of lung tissue to participation in gas
exchange — that is, to an improvement in the
overall ventilation–perfusion ratio. The anatomi-
cal and the functional lung recruitment can co-
incide only if the restoration of aeration of pulmo-
nary units, as detected by CT, occurs in association
with the absence of a change in perfusion of the
same units. Our data support the hypothesis that
anatomical and functional lung recruitment are at
least partially dissociated.
We believe that knowledge of the percentage
of potentially recruitable lung may be important
for establishing the therapeutic efficacy of PEEP.
Setting levels of PEEP independently of the per-
centage of potentially recruitable lung, which was
the strategy used by Brower et al.,12 may offset the
possible benefits of PEEP. Our data show that
the use of higher PEEP levels in patients with a
lower percentage of potentially recruitable lung
provides little benefit and may actually be harm-
ful. To determine whether different levels of PEEP
may affect the outcome among patients with
acute lung injury or ARDS, a formal study will be
necessary, but it should be limited to patients with
a higher percentage of potentially recruitable lung.
Although the use of higher PEEP levels seems
appropriate in these patients, it should be for-
mally tested. Since about 60 percent of lung
parenchyma is already open to aeration in pa-
tients with a higher percentage of potentially re-
cruitable lung, this portion of the lung may be
unnecessarily exposed to increased stress and
strain with the use of higher PEEP levels.27 While
we wait for such a study to be performed, in our
daily practice we limit the use of PEEP levels of
more than 15 cm of water to patients with a high-
er percentage of potentially recruitable lung28 and
of PEEP levels below 10 cm of water to those with
a lower percentage of potentially recruitable lung.
Dr. Gattinoni reports having received consulting and lecture
fees from KCI; Dr. Ranieri reports serving as a consultant to
Maquet and having received grant support from Tyco; and Dr.
Quintel reports having received consulting fees from Siemens/
Maquet, Novalung, Dräger Medical, Abbott Laboratories, and
Gambro. No other potential conflict of interest relevant to this
article was reported.
We are indebted to Angelo Colombo, M.D., Ph.D., of the Tera-
pia Intensiva Neuroscienze, Fondazione Istituto di Ricovero e Cura
a Carattere Scientifico (IRCCS)–Ospedale Maggiore Policlinico,
Mangiagalli, Regina Elena di Milano, Milan, for statistical ad-
vice; to Pietro Biondetti, M.D., Marco Lazzarini, M.D., Benedetta
Finamore, M.D., and Cristian Bonelli of the Dipartimento di
www.nejm.org april 27, 2006 1785
 lung recruitment in patients with the acute respir atory distress syndrome

Radiologia, Fondazione IRCCS–Ospedale Maggiore Policlinico,
Mangiagalli, Regina Elena di Milano, Milan, for technical assis-
tance with analysis of CT findings; to Milena Racagni, M.D.,
Laura Landi, M.D., Alice D’Adda, M.D., Serena Azzari, M.D., Sonia
Terragni, M.D., Federico Polli, M.D., Paola Cozzi, M.D., Giuliana
Motta, M.D., Federica Tallarini, M.D., Cristian Carsenzola, M.D.,
and Monica Chierichetti, M.D., of the Istituto di Anestesiologia
e Rianimazione, Fondazione IRCCS–Ospedale Maggiore Policlin-
ico, Mangiagalli, Regina Elena di Milano, Università degli Studi di
Milano, Milan, for help with the data analysis; to Ferdinando
Raimondi, M.D., of the I Servizio di Anestesia e Rianimazione,
Ospedale Luigi Sacco di Milano, Milan; Antonio Pesenti, M.D., of
the Dipartimento di Medicina Perioperatoria e Terapia Intensiva,
Azienda Ospedaliera S. Gerardo di Monza, Università degli Studi
Milano–Bicocca, Milan; Roberto Fumagalli, M.D., of the Diparti-
mento di Anestesia e Rianimazione, Ospedali Riuniti di Bergamo,
Università degli Studi Milano-Bicocca, Milan; and Danilo Radriz-
zani, M.D., of the Dipartimento Emergenza Urgenza, Ospedale
Civile di Legnano, Legnano, Italy, for their cooperation in retriev-
ing data for control groups; to the study patients for their partici-
pation; and to the physicians and nursing staff of the participat-
ing units for their valuable cooperation.

References
1. Ashbaugh DG, Bigelow DB, Petty TL,
Levine BE. Acute respiratory distress in
adults. Lancet 1967;2:319-23.
2. Ware LB, Matthay MA. The acute re-
spiratory distress syndrome. N Engl J Med
2000;342:1334-49.
3. Dreyfuss D, Saumon G. Ventilator-
induced lung injury: lessons from experi-
mental studies. Am J Respir Crit Care Med
1998;157:294-323.
4. Pontoppidan H, Geffin B, Lowenstein
E. Acute respiratory failure in the adult.
N Engl J Med 1972;287:799-806.
5. Gattinoni L, Agostoni A, Pesenti A, et
al. Treatment of acute respiratory failure
with low-frequency positive-pressure ven-
tilation and extracorporeal removal of
CO2. Lancet 1980;2:292-4.
6. Kolobow T, Moretti MP, Fumagalli R,
et al. Severe impairment in lung function
induced by high peak airway pressure
during mechanical ventilation: an experi-
mental study. Am Rev Respir Dis 1987;
135:312-5.
7. Hickling KG, Henderson SJ, Jackson R.
Low mortality associated with low volume
pressure limited ventilation with permis-
sive hypercapnia in severe adult respira-
tory distress syndrome. Intensive Care Med
1990;16:372-7.
8. Lachmann B. Open up the lung and
keep the lung open. Intensive Care Med
1992;18:319-21.
9. Amato MB, Barbas CS, Medeiros DM,
et al. Effect of a protective-ventilation
strategy on mortality in the acute respira-
tory distress syndrome. N Engl J Med 1998;
338:347-54.
10. Ranieri VM, Suter PM, Tortorella C,
et al. Effect of mechanical ventilation on
inflammatory mediators in patients with
acute respiratory distress syndrome: a ran-
domized controlled trial. JAMA 1999;282:
54-61.
11. Ventilation with lower tidal volumes
as compared with traditional tidal vol-
umes for acute lung injury and the acute
respiratory distress syndrome: the Acute
Respiratory Distress Syndrome Network.
N Engl J Med 2000;342:1301-8.
12. Brower RG, Lanken PN, MacIntyre N,
et al. Higher versus lower positive end-
expiratory pressures in patients with the
acute respiratory distress syndrome. N Engl
J Med 2004;351:327-36.
13. Crotti S, Mascheroni D, Caironi P, et
al. Recruitment and derecruitment during
acute respiratory failure: a clinical study.
Am J Respir Crit Care Med 2001;164:131-
40.
14. Villagra A, Ochagavia A, Vatua S, et al.
Recruitment maneuvers during lung pro-
tective ventilation in acute respiratory dis-
tress syndrome. Am J Respir Crit Care
Med 2002;165:165-70.
15. Grasso S, Mascia L, Del Turco M, et al.
Effects of recruiting maneuvers in patients
with acute respiratory distress syndrome
ventilated with protective ventilatory strat-
egy. Anesthesiology 2002;96:795-802.
16. Brower RG, Morris A, MacIntyre N, et
al. Effects of recruitment maneuvers in
patients with acute lung injury and acute
respiratory distress syndrome ventilated
with high positive end-expiratory pressure.
Crit Care Med 2003;31:2592-7. [Erratum,
Crit Care Med 2004;32:907.]
17. Levy MM. PEEP in ARDS — how much
is enough? N Engl J Med 2004;351:389-91.
18. Gattinoni L, D’Andrea L, Pelosi P, Vi-
tale G, Pesenti A, Fumagalli R. Regional
effects and mechanism of positive end-
expiratory pressure in early adult respiratory
distress syndrome. JAMA 1993;269:2122-7.
[Erratum, JAMA 1993;270:1814.]
19. Gattinoni L, Caironi P, Pelosi P, Good-
man LR. What has computed tomography
taught us about the acute respiratory dis-
tress syndrome? Am J Respir Crit Care
Med 2001;164:1701-11.
20. Bernard GR, Artigas A, Brigham KL,
et al. The American-European Consensus
Conference on ARDS: definitions, mecha-
nisms, relevant outcomes, and clinical
trial coordination. Am J Respir Crit Care
Med 1994;149:818-24.
21. Piacentini E, Villagra A, Lopez-Agui-
lar J, Blanch L. Clinical review: the impli-
cations of experimental and clinical stud-
ies of recruitment maneuvers in acute
lung injury. Crit Care 2004;8:115-21.
22. Lapinsky SE, Mehta S. Bench-to-bed-
side review: recruitment and recruiting
maneuvers. Crit Care 2005;9:60-5.
23. Le Gall JR, Lemeshow S, Saulnier F.
A new Simplified Acute Physiology Score
(SAPS II) based on a European/North Amer-
ican multicenter study. JAMA 1993;270:
2957-63. [Erratum, JAMA 1994;271:1321.]
24. Papadakos PJ, Lachmann B. The open
lung concept of alveolar recruitment can
improve outcome in respiratory failure
and ARDS. Mt Sinai J Med 2002;69:73-7.
25. Pelosi P, D’Andrea L, Vitale G, Pesenti
A, Gattinoni L. Vertical gradient of re-
gional lung inflation in adult respiratory
distress syndrome. Am J Respir Crit Care
Med 1994;149:8-13.
26. Terashima T, Matsubara H, Nakamura
M, et al. Local Pseudomonas instillation
induces contralateral lung injury and
plasma cytokines. Am J Respir Crit Care
Med 1996;153:1600-5.
27. Gattinoni L, Carlesso E, Cadringher P,
Valenza F, Vagginelli F, Chiumello D. Phys-
ical and biological triggers of ventilator-
induced lung injury and its prevention.
Eur Respir J Suppl 2003;47:15s-25s.
28. Grasso S, Fanelli V, Cafarelli A, et al.
Effects of high versus low positive end-
expiratory pressures in acute respiratory
distress syndrome. Am J Respir Crit Care
Med 2005;171:1002-8.
Copyright © 2006 Massachusetts Medical Society.

1786 n engl j med 354;17 www.nejm.org april 27, 2006

The New England Journal of Medicine

Downloaded from nejm.org on August 15, 2014. For personal use only. No other uses without permission.
Copyright © 2006 Massachusetts Medical Society. All rights reserved.